Adaptive Recall is a persistent memory system for AI assistants built by AI Apps API. It offers eight API tools: store, recall, update, forget, graph, status, snapshot, and feedback. The system operates over MCP for Claude Code and other CLI tools, and via plain HTTP REST for any application. Authentication uses bearer tokens, and JSON is used for input and output. Users receive 500 memories free with no credit card required.

Adaptive Recall employs four retrieval strategies that run in parallel: vector similarity, temporal recency, full-text keyword, and knowledge graph traversal. It uses ACT‑R cognitive scoring from 30 years of research to rank results. The system learns which strategy to prioritize for each query type. Automatic construction of a knowledge graph extracts entities and relationships from stored memories, creating pathways for retrieval beyond text similarity.

Memories are not static; they progress through stages, gaining or losing confidence based on corroborating evidence and fading when not accessed. Adaptive Recall trains machine‑learning models on usage patterns and validates every parameter change against real query history. It monitors its own retrieval quality and performs curiosity‑driven knowledge‑gap detection. The API provides tools to update or forget memories and to retrieve graph or status snapshots.